A continuous casting platform ladle hoisting state detection system and method
Patent Information
- Application Number
- CN202611076462.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-20
- Publication Date
- 2026-10-27
AI Technical Summary
传统作业模式存在以下缺陷：工人需近距离观察数百吨高温钢包，面临钢水喷溅、设备碰撞等安全威胁；高温热浪导致视觉畸变、粉尘水汽遮挡视线，人工误判率较高；单次状态确认需3~5分钟，导致后续机器人作业待机时间延长，连铸产线整体效率损失15%以上
[0015]本发明的有益效果在于：本发明采用机器视觉技术在线检测钢包、钢包耳轴、行车吊钩信息，建立图像检测信息状态与时序结合的判断准则，智能判定钢包吊运状态，无人化给连铸平台区域装备提供钢包吊运状态信息，有利于提高钢铁冶金智能化水平，减轻工人劳动强度。方法简单易施，实用性和适用性较大。
Smart Images
Figure CN122989576A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of industrial automation technology and relates to a system and method for detecting the status of ladle hoisting on a continuous casting platform. Background Technology
[0002] In the iron and steel metallurgical industry, continuous casting, as a core process connecting steelmaking and rolling, directly impacts production efficiency and safety due to its level of automation. The ladle receiving area of the continuous casting platform is a critical area for ladle hoisting, positioning, and pouring operations. In recent years, with the development of intelligent manufacturing technology, unmanned operating equipment such as disassembling and assembling sliding plate cylinder robots and automatic temperature measurement and sampling robots have been gradually introduced into the ladle receiving area. However, accurate detection of the ladle hoisting status remains the core bottleneck restricting the unmanned operation of the entire process.
[0003] Currently, steel ladles are lifted by casting cranes using hooks to suspend the ladle trunnions. However, the confirmation of their positioning, alignment, and hook / unhooking status still relies on manual visual judgment on-site. This traditional operating mode has the following drawbacks: workers need to observe hundreds of tons of high-temperature steel ladles at close range, facing safety threats such as molten steel splashes and equipment collisions; high temperatures cause visual distortion, and dust and moisture obstruct vision, resulting in a high rate of human error; a single status confirmation takes 3-5 minutes, leading to extended standby time for subsequent robot operations and an overall efficiency loss of over 15% for the continuous casting production line.
[0004] Existing technologies attempt to use physical sensors for condition detection, but none are suitable for the extreme working conditions of continuous casting platforms. Contact sensors are prone to failure under high-temperature radiation, and the impact force of the ladle falling causes a sharp drop in the life of the mechanical structure; non-contact sensors are significantly affected by moisture and iron oxide dust, resulting in a high false trigger rate; video monitoring systems only achieve remote visualization and still require manual judgment, failing to solve the problem of automated decision-making.
[0005] Although machine vision is widely used in industrial inspection, the special working conditions of the continuous casting ladle receiving position cause existing technologies to fail: the ladle trunnion and hook structure is complex, and the surface is covered with oxide scale and molten steel splashes, making traditional edge detection algorithms ineffective; high temperature heat waves cause sudden changes in air refractive index, resulting in image distortion; the light intensity in the continuous casting workshop fluctuates from 100 to 10000 Lux, making it difficult for conventional image enhancement algorithms to work stably; it is necessary to simultaneously determine multiple dimensions of parameters such as the stability of ladle placement, trunnion-hook engagement depth, and hook opening and closing status, and existing algorithms lack the ability to make decisions based on multi-feature fusion.
[0006] Therefore, in order to meet the demand for unmanned operation of ladle receiving position on continuous casting platform, it is urgent to develop a ladle hoisting status detection system and method that is resistant to environmental interference and has multi-state collaborative discrimination, to break through the technical barriers of accurate extraction of visual features and intelligent status decision-making under harsh working conditions, and realize fully automated closed-loop control of hoisting process. Summary of the Invention
[0007] In view of this, the purpose of the present invention is to provide a system and method for detecting the hoisting status of steel ladles on a continuous casting platform.
[0008] To achieve the above objectives, the present invention provides the following technical solution: A continuous casting platform ladle hoisting status detection system includes an industrial camera and a computing terminal, wherein the computing terminal is equipped with an image processing module and a hoisting status judgment module; The industrial camera is installed in the ladle receiving area of the continuous casting platform to collect image data during ladle hoisting and input it into the computing terminal; The computing terminal analyzes and processes the image data collected by the industrial camera through the image processing module, and detects the status of the ladle, ladle trunnion and crane hook; then it judges and outputs the ladle hoisting status information through the hoisting status judgment module.
[0009] Furthermore, the image processing module is used to analyze and process the image data acquired by the industrial camera, and to detect the state of the ladle, ladle trunnion, and crane hook using machine vision technology. The machine vision technology uses the YOLO lightweight object detection neural network framework, and uses the image data of the ladle, ladle trunnion, and crane hook for model training, so that the network input is 2D image data, and the network output is the rectangular bounding box information of the ladle, ladle trunnion, and crane hook in the image. The rectangular bounding box information includes the center coordinates (x, y), width W, and height H of the rectangle.
[0010] Furthermore, the hoisting status judgment module is used to judge the hoisting status of the trolley on the ladle. It continuously receives the rectangular bounding box information output by the image processing module and outputs the ladle hoisting status information.
[0011] Furthermore, the hoisting status information includes hoisting to the large package slewing platform, moving away from the large package slewing platform, the crane hook being unhooked, and the crane hook being engaged.
[0012] Furthermore, in the hoisting status judgment module, the judgment criterion combining the image detection information status and timing is as follows: When the image changes from no ladle information to ladle information, it is determined that the crane is lifting the ladle to the ladle turntable. At this time, the IoU of the ladle trunnion and the crane hook changes slightly, and the lateral distance D = |x1-x2| between their center coordinates changes slightly. When the IoU of the ladle trunnion and the crane hook shows a gradually decreasing trend, and the lateral distance D between their center coordinates shows a gradually increasing trend, it is determined that the crane has completed the work of hoisting the ladle to the ladle slewing platform and is performing the unhooking and removal operation. When D > W1 + W2 + K, it is determined that the crane has completed uncoupling and is moving away from the turntable area; When the image continuously shows information about the ladle and ladle trunnion but no information about the overhead crane hook, and the x-coordinate of the ladle and ladle trunnion changes only slightly, it is determined that the ladle is in a state where it is waiting to be lifted off the ladle turntable by the overhead crane. When the crane hook is detected in the image, it is determined that the crane has started the ladle removal operation; When the lateral distance D between the center coordinates of the ladle trunnion and the crane hook gradually decreases, it is determined that the crane hook is approaching the ladle trunnion. When the lateral distance D between the center coordinates of the ladle trunnion and the overhead crane hook changes slightly, and the IoU between the two changes slightly, it is determined that the overhead crane hook has been attached to the ladle trunnion, and the overhead crane will perform the operation of removing the ladle. When the detection information for the ladle, ladle trunnion, and crane hook in the image changes from "all present" to "never present," the crane has completed the ladle removal operation.
[0013] On the other hand, the present invention provides a method for detecting the ladle hoisting status of a continuous casting platform, comprising the following steps: S1: Images are captured by an industrial camera in the ladle receiving area of the continuous casting platform. The shooting range of the industrial camera includes the ladle and the area of the ladle slewing arm. S2: The computing terminal collects images captured by the industrial camera, analyzes and processes them through the image processing module, and detects the status of the ladle, ladle trunnion and overhead crane hook; S3: The hoisting status judgment module judges and outputs the hoisting status information of the steel ladle based on the judgment criteria that combine the image detection information status and timing.
[0014] Furthermore, it also includes S4: the equipment in the continuous casting platform area further performs various operations based on the ladle hoisting status.
[0015] The beneficial effects of this invention are as follows: This invention uses machine vision technology to detect information about ladles, ladle trunnions, and crane hooks online, establishes a judgment criterion combining the status of image detection information with time sequence, intelligently determines the ladle hoisting status, and provides unmanned ladle hoisting status information to the equipment in the continuous casting platform area, which is conducive to improving the level of intelligence in iron and steel metallurgy and reducing the labor intensity of workers. The method is simple to implement and has wide practicality and applicability.
[0016] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description
[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 Structural diagram of the ladle hoisting status monitoring system for continuous casting platform; Figure 2 Flowchart of the method for detecting the hoisting status of steel ladles on a continuous casting platform. Detailed Implementation
[0018] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.
[0019] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.
[0020] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.
[0021] Example 1: like Figure 1 As shown, this invention provides a continuous casting platform ladle hoisting status detection system, including an industrial camera, an image processing module, a hoisting status judgment module, and a computing terminal. The industrial camera is fixedly installed in the ladle receiving area of the continuous casting platform to collect image data during ladle hoisting. Preferably, the industrial camera's shooting range includes the ladle and the area of the ladle's rotating arm. The computing terminal is the core logic processing area; the industrial camera is communicatively connected to the computing terminal, and both the image processing module and the hoisting status judgment module are mounted on the computing terminal.
[0022] The image processing module is responsible for analyzing and processing the image data acquired by the industrial camera, and using machine vision technology to detect the state of the ladle trunnion and the crane hook. Specifically: First, the YOLO series object detection neural network framework is used to acquire image data of the ladle, ladle trunnion, and crane hook for model training, so that the network input is 2D image data, and the network output is the bounding rectangle of the ladle, ladle trunnion, and crane hook in the image. The bounding box information includes the center coordinates (x, y), width (W), and height (H) of the rectangle.
[0023] In application, the image data acquired by the aforementioned industrial camera is input into the image processing module, and the image processing module outputs the bounding rectangle information of the steel ladle, steel ladle trunnion, and crane hook it detected.
[0024] The hoisting status judgment module is used to judge the hoisting status of the trolley on the ladle. It continuously receives the boundary rectangle information output by the image processing module and outputs the ladle hoisting status information, including hoisting to the ladle slewing table, moving away from the ladle slewing table, the trolley hook being unhooked, and the trolley hook being engaged.
[0025] Specifically, the hoisting status judgment module establishes a judgment criterion that combines the image detection information status with the timing. When the image changes from no ladle information to ladle information, it is judged that the crane is hoisting the ladle to the ladle turntable. At this time, the IoU (Intersection over Union) of the ladle trunnion and the crane hook changes slightly, and the lateral distance D (D = |x1-x2|) between their center coordinates changes slightly. When the IoU of the ladle trunnion and the crane hook changes significantly and gradually decreases, and the lateral distance D between their center coordinates changes significantly and gradually increases, it is judged that the crane has completed the work of hoisting the ladle to the ladle turntable and is performing the unhooking and removal operation. When D > W1 + W2 + K, it is judged that the crane has completed the unhooking and is moving away from the turntable area. Preferably, K is a fixed value, set according to the pixel value of the camera image, usually in the range of 20-100. When the image continuously shows information about the ladle and ladle trunnions but no information about the overhead crane hook, and the x-coordinates of the ladle and ladle trunnions change only slightly, the ladle is in the state of waiting to be lifted off the ladle turntable by the overhead crane. When the overhead crane hook is detected in the image, the overhead crane begins the ladle lifting operation. When the lateral distance D between the center coordinates of the ladle trunnions and the overhead crane hook gradually decreases, the overhead crane hook is approaching the ladle trunnions. When the lateral distance D between the center coordinates of the ladle trunnions and the overhead crane hook changes only slightly, and the IoU between the two changes only slightly, the overhead crane hook has already hooked onto the ladle trunnions, and the overhead crane performs the ladle removal operation. When the detection information of the ladle, ladle trunnions, and overhead crane hook in the image changes from all being present to continuously being absent, the overhead crane has completed the ladle removal operation.
[0026] In application, the boundary rectangle information output by the above image processing module is input into the hoisting status judgment module. Based on the judgment criteria that combine the image detection information status and timing, the ladle hoisting status output is obtained: hoisted to the ladle slewing table, moved away from the ladle slewing table, the crane hook has been unhooked, and the crane hook has been attached.
[0027] The computing terminal is the core logic processing area. The industrial camera is communicatively connected to the computing terminal. The image processing module and the hoisting status judgment module are both mounted on the computing terminal.
[0028] Example 2: like Figure 2 As shown in the figure, this embodiment provides a method for detecting the hoisting status of a ladle on a continuous casting platform, including the following steps: Step S1: Install an industrial camera in the ladle receiving area of the continuous casting platform. The shooting range of the industrial camera includes the ladle and the area of the ladle slewing arm.
[0029] Step S2: Build an image processing module based on machine vision technology.
[0030] Specifically, the YOLO series object detection neural network framework is used. During production operations, image data of the ladle, ladle trunnion, and crane hook are acquired using an industrial camera for model training. The input to the model network is 2D image data, and the output of the model network is the bounding rectangle of the ladle, ladle trunnion, and crane hook in the image. The bounding box information includes the center coordinates (x, y), width (W), and height (H) of the rectangle. Step S3: Based on the judgment criteria that combine image detection information status and timing, a hoisting status judgment module is built.
[0031] Specifically, when the image transitions from having no ladle information to having ladle information, it is determined that the crane is in the working state of lifting the ladle to the ladle turntable. At this time, the IoU (Intersection over Union) of the ladle trunnion and the crane hook changes slightly, and the lateral distance D (D = |x1-x2|) between their center coordinates changes slightly. When the IoU of the ladle trunnion and the crane hook changes significantly and gradually decreases, and the lateral distance D between their center coordinates changes significantly and gradually increases, it is determined that the crane has completed the work of lifting the ladle to the ladle turntable and is performing the unhooking and removal operation. When D > W1 + W2 + K, it is determined that the crane has completed the unhooking and is moving away from the turntable area. Preferably, K is a fixed value, set according to the pixel value of the camera image, usually in the range of 20-100.
[0032] When the image continuously shows information about the ladle and ladle trunnions but no information about the overhead crane hook, and the x-coordinates of the ladle and ladle trunnions change only slightly, the ladle is in the state of waiting to be lifted off the ladle turntable by the overhead crane. When the overhead crane hook is detected in the image, the overhead crane begins the ladle lifting operation. When the lateral distance D between the center coordinates of the ladle trunnions and the overhead crane hook gradually decreases, the overhead crane hook is approaching the ladle trunnions. When the lateral distance D between the center coordinates of the ladle trunnions and the overhead crane hook changes only slightly, and the IoU between the two changes only slightly, the overhead crane hook has already hooked onto the ladle trunnions, and the overhead crane performs the ladle removal operation. When the detection information of the ladle, ladle trunnions, and overhead crane hook in the image changes from all being present to continuously being absent, the overhead crane has completed the ladle removal operation.
[0033] Step S4: The computing terminal establishes a communication connection with the industrial camera and is equipped with an image processing module and a hoisting status judgment module.
[0034] Step S5, system production use: The image processing module uses the established target detection model to process the images acquired in real time by the industrial camera, obtains the information of the ladle, ladle trunnion and crane hook in the current image, and outputs the information to the hoisting status judgment module.
[0035] The hoisting status judgment module outputs the hoisting status of the ladle based on the pre-set judgment criteria that combine the image detection information status and timing: hoisted to the ladle slewing platform, moved away from the ladle slewing platform, the crane hook has been unhooked, and the crane hook has been engaged.
[0036] The equipment in the continuous casting platform area can further perform various operations based on the ladle hoisting status.
[0037] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A continuous casting platform ladle hoisting status detection system, characterized in that: It includes an industrial camera and a computing terminal, wherein the computing terminal is equipped with an image processing module and a hoisting status judgment module; The industrial camera is installed in the ladle receiving area of the continuous casting platform to collect image data during ladle hoisting and input it into the computing terminal; The computing terminal analyzes and processes the image data collected by the industrial camera through the image processing module, and detects the status of the ladle, ladle trunnion and crane hook; then it judges and outputs the ladle hoisting status information through the hoisting status judgment module.
2. The continuous casting platform ladle hoisting status detection system according to claim 1, characterized in that: The image processing module is used to analyze and process the image data acquired by the industrial camera, and uses machine vision technology to detect the state of the ladle, ladle trunnion, and crane hook. The machine vision technology adopts the YOLO lightweight object detection neural network framework, and uses the image data of the ladle, ladle trunnion, and crane hook for model training, so that the network input is 2D image data, and the network output is the rectangular bounding box information of the ladle, ladle trunnion, and crane hook in the image. The rectangular bounding box information includes the center coordinates (x, y), width W, and height H of the rectangle.
3. The continuous casting platform ladle hoisting status detection system according to claim 1, characterized in that: The hoisting status judgment module is used to judge the hoisting status of the steel ladle by the crane. It continuously receives the rectangular bounding box information output by the image processing module and outputs the steel ladle hoisting status information.
4. The continuous casting platform ladle hoisting status detection system according to claim 1, characterized in that: The hoisting status information includes hoisting to the large package slewing platform, moving away from the large package slewing platform, the crane hook being unhooked, and the crane hook being engaged.
5. The continuous casting platform ladle hoisting status detection system according to claim 1, characterized in that: In the hoisting status judgment module, the judgment criterion combining the image detection information status and timing is as follows: When the image changes from no ladle information to ladle information, it is determined that the crane is lifting the ladle to the ladle turntable. At this time, the IoU of the ladle trunnion and the crane hook changes slightly, and the lateral distance D = |x1-x2| between their center coordinates changes slightly. When the IoU of the ladle trunnion and the crane hook shows a gradually decreasing trend, and the lateral distance D between their center coordinates shows a gradually increasing trend, it is determined that the crane has completed the work of hoisting the ladle to the ladle slewing platform and is performing the unhooking and removal operation. When D > W1 + W2 + K, it is determined that the crane has completed uncoupling and is moving away from the turntable area; When the image continuously shows information about the ladle and ladle trunnion but no information about the overhead crane hook, and the x-coordinate of the ladle and ladle trunnion changes only slightly, it is determined that the ladle is in a state where it is waiting to be lifted off the ladle turntable by the overhead crane. When the crane hook is detected in the image, it is determined that the crane has started the ladle removal operation; When the lateral distance D between the center coordinates of the ladle trunnion and the crane hook gradually decreases, it is determined that the crane hook is approaching the ladle trunnion. When the lateral distance D between the center coordinates of the ladle trunnion and the overhead crane hook changes slightly, and the IoU between the two changes slightly, it is determined that the overhead crane hook has been attached to the ladle trunnion, and the overhead crane will perform the operation of removing the ladle. When the detection information for the ladle, ladle trunnion, and crane hook in the image changes from "all present" to "never present," the crane has completed the ladle removal operation.
6. A method for detecting the hoisting status of a steel ladle on a continuous casting platform, characterized in that: Includes the following steps: S1: Images are captured by an industrial camera in the ladle receiving area of the continuous casting platform. The shooting range of the industrial camera includes the ladle and the area of the ladle slewing arm. S2: The computing terminal collects images captured by the industrial camera, analyzes and processes them through the image processing module, and detects the status of the ladle, ladle trunnion and overhead crane hook; S3: The hoisting status judgment module judges and outputs the hoisting status information of the steel ladle based on the judgment criteria that combine the image detection information status and timing.
7. The method for detecting the ladle hoisting status of a continuous casting platform according to claim 6, characterized in that: It also includes S4: the equipment in the continuous casting platform area further performs various operations based on the ladle hoisting status.